Recent Progress in Approximation Algorithms for the Traveling Salesman Problem
|
|
- Abel Jefferson
- 6 years ago
- Views:
Transcription
1 Recent Progress in Approximation Algorithms for the Traveling Salesman Problem Lecture 4: s-t path TSP for graph TSP David P. Williamson Cornell University July 18-22, 2016 São Paulo School of Advanced Science on Algorithms, Combinatorics, and Optimization
2 s-t path TSP Recall the s-t path TSP: Usual TSP input plus s, t V, find a min-cost path from s to t visiting all other nodes in between (an s-t Hamiltonian path).
3 A Linear Programming Relaxation Min e E c e x e { 1, v = s, t, subject to: x(δ(v)) = 2, v s, t, { 1, S {s, t} = 1, x(δ(s)) 2, S {s, t} 1, 0 x e 1, e E, where δ(s) is the set of edges with exactly one endpoint in S, and x(e ) e E x e. Lemma Any solution x feasible for the s-t path TSP LP relaxation is in the spanning tree polytope.
4 Best-of-Many Christofides Algorithm An, Kleinberg, Shmoys (2012) propose the Best-of-Many Christofides algorithm: given optimal LP solution x, compute convex combination of spanning trees k x = λ i χ Fi. i=1 For each spanning tree F i, let T i be the set of vertices whose parity needs fixing, let J i be the minimum-cost T i -join. Find s-t Hamiltonian path by shortcutting F i J i. Return the shortest path found over all i.
5 From last time To prove that Best-of-Many Christofides is at most 5 3 OPT LP for optimal LP solution x, show that is feasible for the T i -join LP: y i = 1 3 χ F i x Min e E c e x e subject to: x(δ(s)) 1, S V, S T i odd x e 0, e E.
6 Improvement? To do better, we need to improve the analysis for the costs of the T i -joins; recall that we use that is feasible for the T i -join LP. Consider y i = 1 3 χ F i x y i = αχ Fi + βx. Then the cost of the best s-t Hamiltonian path is at most (1 + α + β)opt LP.
7 Improvement? Proof that y i feasible for T i -join LP had two cases. Assume S odd ( S T i odd). If S {s, t} 1, then y i (δ(s)) = α F i δ(s) + βx (δ(s)) α + 2β. We will want α + 2β 1, so the T i -join LP constraint is satisfied.
8 Improvement? If S {s, t} = 1, then y i (δ(s)) = α F i δ(s) + βx (δ(s)) 2α + βx (δ(s)). Since we assume α + 2β 1, we only run into problems if x (δ(s)) < 1 2α. β Note that α = 0, β = 1 2 works if x (δ(s)) 2 for all S V, and gives a tour of cost at most 3 2 OPT LP.
9 Improvement? If S {s, t} = 1, then y i (δ(s)) = α F i δ(s) + βx (δ(s)) 2α + βx (δ(s)). Since we assume α + 2β 1, we only run into problems if x (δ(s)) < 1 2α. β Note that α = 0, β = 1 2 works if x (δ(s)) 2 for all S V, and gives a tour of cost at most 3 2 OPT LP. So focus on s-t cuts for which x (δ(s)) < 2, and add an extra correction term to y i to handle these cuts.
10 τ-narrow Cuts Definition S is τ-narrow if x (δ(s)) < 1 + τ for fixed τ 1. Only S such that S {s, t} = 1 are τ-narrow. Definition Let C τ be all τ-narrow cuts S with s S.
11 τ-narrow Cuts The τ-narrow cuts in C τ have a nice structure. Theorem (An, Kleinberg, Shmoys (2012)) If S 1, S 2 C τ, S 1 S 2, then either S 1 S 2 or S 2 S 1. So the τ-narrow cuts look like s Q 1 Q 2 Q k V. s... t
12 Correction Factor Let e Q be the minimum-cost edge in δ(q). Then consider the following (from Gao (2014)): y i = αχ Fi + βx + (1 2α βx (δ(q))) χ eq Q C τ, Q T i odd for α, β, τ 0 such that α + 2β = 1 and τ = 1 2α β 1. Theorem y i is feasible for the T i -join LP.
13 Parameters By choosing α = 1 2 5, β = 1 5, τ = 3 5, then one can show that the total cost is at most OPT LP. 2
14 Today Today: Combine the two special cases: Look at s-t path TSP in the case of graph TSP instances (e.g. input is undirected graph, cost c(i, j) is number of edges in shortest i-j path)
15 Integrality Gap The performance of Best-of-Many Christofides cannot do better than the integrality gap of the LP relaxation.
16 Integrality Gap The performance of Best-of-Many Christofides cannot do better than the integrality gap of the LP relaxation. The integrality gap is µ sup OPT OPT LP over all instances of the problem.
17 Integrality Gap The performance of Best-of-Many Christofides cannot do better than the integrality gap of the LP relaxation. The integrality gap is over all instances of the problem. µ sup OPT OPT LP Note that An, Kleinberg, Shmoys have shown µ 1.618, since Best-of-Many Christofides algorithm finds a tour of cost at most OPT LP.
18 Integrality Gap We can show a lower bound on the integrality gap using an instance of graph TSP: input is a graph G = (V, E), cost c e for e = (i, j) is number of edges in a shortest i-j path in G. s t k
19 Integrality Gap We can show a lower bound on the integrality gap using an instance of graph TSP: input is a graph G = (V, E), cost c e for e = (i, j) is number of edges in a shortest i-j path in G. s t k OPT LP 2k
20 Integrality Gap We can show a lower bound on the integrality gap using an instance of graph TSP: input is a graph G = (V, E), cost c e for e = (i, j) is number of edges in a shortest i-j path in G. s t k OPT 3k
21 Integrality Gap We can show a lower bound on the integrality gap using an instance of graph TSP: input is a graph G = (V, E), cost c e for e = (i, j) is number of edges in a shortest i-j path in G. s t k OPT OPT LP 3 2 as k
22 Graph Instances Sebő and Vygen (2014) show that for graph TSP instances of s-t path TSP, can get a 3 2-approximation algorithm (i.e. the algorithm produces a solution of cost at most 3 2 OPT LP), so the integrality gap is tight for these instances. We ll present a simplified version of this result due to Gao (2013).
23 Graph Instances Given the input graph G = (V, E) and an optimal solution, can replace any edge (i, j) in the optimal solution with the i-j path in G since these have the same cost. So finding an optimal solution is equivalent to finding a multiset F of edges such that (V, F ) is connected, deg F (s) and deg F (t) are odd, deg F (v) is even for all v V {s, t}, and F is minimum.
24 LP Relaxation Min e E x(e) subject to: x(δ(s)) { 1, S {s, t} = 1, 2, S {s, t} 1, x(e) 0, e E. Let x be an optimal LP solution.
25 Narrow Cuts As before, focus on narrow cuts S such that x (δ(s)) < 2 (i.e. a τ-narrow cut for τ = 1). Recall: Theorem (An, Kleinberg, Shmoys (2012)) If S 1, S 2 are narrow cuts, S 1 S 2, then either S 1 S 2 or S 2 S 1.
26 Proof First need to show that x (δ(s 1 )) + x (δ(s 2 )) x (δ(s 1 S 2 )) + x (δ(s 2 S 1 )). S 1 S 2
27 Proof Theorem (An, Kleinberg, Shmoys (2012)) If S 1, S 2 are narrow cuts, S 1 S 2, then either S 1 S 2 or S 2 S 1.
28 Narrow Cuts Theorem (An, Kleinberg, Shmoys (2012)) If S 1, S 2 are narrow cuts, S 1 S 2, then either S 1 S 2 or S 2 S 1. So the narrow cuts look like s S 1 S 2 S k V. s... t Let S 0, S k+1 V, L i S i S i 1.
29 Key Idea Find a tree spanning L i in the support of x for each i. Connect each of these via a single edge from L i to L i+1. Let F be the resulting tree, T the vertices in F whose parity needs changing. Then F = n 1 and δ(s i ) F = 1 for each narrow cut S i. s L 1 L 2... L k t
30 Key Lemma Recall: Lemma Let S be an odd set. If S {s, t} = 1, then F δ(s) is even. Min e E c(e)x(e) subject to: x(δ(s)) 1, S V, S T odd x(e) 0, e E. Lemma y = 1 2 x is feasible for the the T -join LP.
31 Proof
32 Gao (2013) Theorem (Gao (2013)) For spanning tree F constructed by the algorithm, let J be a minimum-cost T -join. Then c(f J) 3 2 OPT LP. Min e E x(e) subject to: x(δ(s)) { 1, S {s, t} = 1, 2, S {s, t} 1, x(e) 0, e E.
33 Proof
34 Last Lemma Let E(x ) = {e E : x (e) > 0} be the support of LP solution x, H = (V, E(x )) the support graph of x, H(S) the graph induced by a set S of vertices. Lemma (Gao (2013)) ( ) For 1 p q k + 1, H p i q L i is connected. s L 1 L 2... L k t
35 Proof
36 One Big Question Is there a 3 2-approx. alg. for s-t path TSP for general costs?
37 One Idea Idea: Construct a spanning tree F just as in Gao s algorithm for the graph case. Then again y = 1 2 x is feasible for the T -join LP, and the overall cost of F plus the T -join is at most c(f ) e E c(e)x (e).
38 One Idea Idea: Construct a spanning tree F just as in Gao s algorithm for the graph case. Then again y = 1 2 x is feasible for the T -join LP, and the overall cost of F plus the T -join is at most c(f ) e E c(e)x (e). Problem: Not clear how to bound the cost of F. Gao (2014) has an example showing that F can have cost greater than OPT LP.
39 Further directions Best-of-Many Christofides from An et al. works with any possible decomposition of the LP solution into spanning trees. Recent improvements of Vygen (2015), Gottschalk and Vygen (2015), and Sebő and Van Zuylen (2016) all use decompositions that have particular properties.
40 Integrality Gap for standard TSP As with s-t path TSP, we can show a lower bound on the integrality gap using an instance of graph TSP. k
41 Integrality Gap for standard TSP As with s-t path TSP, we can show a lower bound on the integrality gap using an instance of graph TSP. k OPT LP 3k
42 Integrality Gap for standard TSP As with s-t path TSP, we can show a lower bound on the integrality gap using an instance of graph TSP. k OPT 4k
43 Integrality Gap for standard TSP As with s-t path TSP, we can show a lower bound on the integrality gap using an instance of graph TSP. k OPT OPT LP 4 3 as k
44 What about the standard TSP? One can define Best-of-Many Christofides for the standard TSP: solve the subtour LP, get LP solution x. Then since n 1 n x is in the spanning tree polytope, find a decomposition of x into a convex combination of spanning trees. Run Christofides algorithm on each one, return the best solution found.
45 Best-of-Many Christofides Unfortunately, the following example (due to Schalekamp and Van Zuylen) shows that an arbitrary decomposition into spanning trees will not improve on Christofides 3 2-approximation algorithm
46 Experiments Together with an undergraduate (Kyle Genova), we tried several different algorithms for decomposing the subtour LP into spanning trees. We ran these algorithms on several types of instances: 59 Euclidean TSPLIB (Reinelt 1991) instances up to 2103 vertices (avg. 524); 5 non-euclidean TSPLIB instances (gr120, si175, si535, pa561, si1032); 39 Euclidean VLSI instances (Rohe) up to 3694 vertices (avg. 1473); 9 graph TSP instances (Kunegis 2013) up to 1615 vertices (avg. 363).
47 The results Std ColGen ColGen+SR Best Ave Best Ave TSPLIB (E) 9.56% 4.03% 6.44% 3.45% 6.24% VLSI 9.73% 7.00% 8.51% 6.40% 8.33% TSPLIB (N) 5.40% 2.73% 4.41% 2.22% 4.08% Graph 12.43% 0.57% 1.37% 0.39% 1.29% MaxEnt Split Split+SR Best Ave Best Ave Best Ave TSPLIB (E) 3.19% 6.12% 5.23% 6.27% 3.60% 6.02% VLSI 5.47% 7.61% 6.60% 7.64% 5.48% 7.52% TSPLIB (N) 2.12% 3.99% 2.92% 3.77% 1.99% 3.82% Graph 0.31% 1.23% 0.88% 1.77% 0.33% 1.20% Costs given as percentages in excess of the cost of the optimal tour.
48 The results Standard Christofides MST (Rohe VLSI instance XQF131) Splitting off + SwapRound
49 The results BoMC yields more vertices in the tree of degree two.
50 The results So while the tree costs more (as percentage of optimal tour)... Std BOM TSPLIB (E) 87.47% 98.57% VLSI 89.85% 98.84% TSPLIB (N) 92.97% 99.36% Graph 79.10% 98.23%
51 The results...the matching costs much less. Std CG CG+SR MaxE Split Sp+SR TSPLIB (E) 31.25% 11.43% 11.03% 10.75% 10.65% 10.41% VLSI 29.98% 14.30% 14.11% 12.76% 12.78% 12.70% TSPLIB (N) 24.15% 9.67% 9.36% 8.75% 8.77% 8.56% Graph 39.31% 5.20% 4.84% 4.66% 4.34% 4.49%
52 Conclusion Q: Are there empirical reasons to think BoMC might be provably better than Christofides algorithm?
53 Conclusion Q: Are there empirical reasons to think BoMC might be provably better than Christofides algorithm? A: Yes.
54 The Big Open Questions Beat 3 2 for TSP Achieve 3 2 for s-t path TSP Achieve 4 3 for graph TSP
55 The End
Improving Christofides Algorithm for the s-t Path TSP
Cornell University May 21, 2012 Joint work with Bobby Kleinberg and David Shmoys Metric TSP Metric (circuit) TSP Given a weighted graph G = (V, E) (c : E R + ), find a minimum Hamiltonian circuit Triangle
More informationarxiv: v3 [cs.dm] 30 Oct 2017
Layers and Matroids for the Traveling Salesman s Paths Frans Schalekamp 1, András Sebő 2, Vera Traub 3, and Anke van Zuylen 4 1 School of Operations Research and Information Engineering, Department of
More informationThe Salesman s Improved Paths: A 3/2+1/34 Approximation
The Salesman s Improved Paths: A /+1/4 Approximation András Sebő Optimisation Combinatoire (G-SCOP) CNRS, Univ. Grenoble Alpes Grenoble, France andras.sebo@grenoble-inp.fr Anke van Zuylen Department of
More informationJens Vygen. New approximation algorithms for the TSP
Jens Vygen New approximation algorithms for the TSP The traveling salesman problem (TSP) is probably the best known combinatorial optimization problem. Although studied intensively for sixty years, the
More informationThe traveling salesman problem
Chapter 58 The traveling salesman problem The traveling salesman problem (TSP) asks for a shortest Hamiltonian circuit in a graph. It belongs to the most seductive problems in combinatorial optimization,
More informationA Randomized Rounding Approach to the Traveling Salesman Problem
A Randomized Rounding Approach to the Traveling Salesman Problem Shayan Oveis Gharan Amin Saberi. Mohit Singh. Abstract For some positive constant ɛ 0, we give a ( 3 2 ɛ 0)-approximation algorithm for
More informationarxiv: v3 [cs.ds] 21 Jun 2018
Shorter tours and longer detours: Uniform covers and a bit beyond Arash Haddadan, Alantha Newman, R. Ravi arxiv:1707.05387v3 [cs.ds] 21 Jun 2018 June 25, 2018 Abstract Motivated by the well known four-thirds
More informationImproved Approximations for Cubic Bipartite and Cubic TSP
Improved Approximations for Cubic Bipartite and Cubic TSP Anke van Zuylen Department of Mathematics The College of William and Mary, Williamsburg, VA, 23185, USA anke@wm.edu Abstract We show improved approximation
More informationABHELSINKI UNIVERSITY OF TECHNOLOGY
Approximation Algorithms Seminar 1 Set Cover, Steiner Tree and TSP Siert Wieringa siert.wieringa@tkk.fi Approximation Algorithms Seminar 1 1/27 Contents Approximation algorithms for: Set Cover Steiner
More informationIntroduction to Mathematical Programming IE406. Lecture 21. Dr. Ted Ralphs
Introduction to Mathematical Programming IE406 Lecture 21 Dr. Ted Ralphs IE406 Lecture 21 1 Reading for This Lecture Bertsimas Sections 10.2, 10.3, 11.1, 11.2 IE406 Lecture 21 2 Branch and Bound Branch
More informationThe Traveling Salesman Problem: An Overview. David P. Williamson, Cornell University Ebay Research January 21, 2014
The Traveling Salesman Problem: An Overview David P. Williamson, Cornell University Ebay Research January 21, 2014 (Cook 2012) A highly readable introduction Some terminology (imprecise) Problem Traditional
More informationarxiv: v2 [cs.dm] 3 Jun 2018
A 1.5-Approximation for Path TSP Rico Zenklusen * June 5, 2018 arxiv:1805.04131v2 [cs.dm] 3 Jun 2018 Abstract We present a 1.5-approximation for the Metric Path Traveling Salesman Problem (path TSP). All
More informationA simple LP relaxation for the Asymmetric Traveling Salesman Problem
A simple LP relaxation for the Asymmetric Traveling Salesman Problem Thành Nguyen Cornell University, Center for Applies Mathematics 657 Rhodes Hall, Ithaca, NY, 14853,USA thanh@cs.cornell.edu Abstract.
More informationBounds on the Traveling Salesman Problem
Bounds on the Traveling Salesman Problem Sean Zachary Roberson Texas A&M University MATH 613, Graph Theory A common routing problem is as follows: given a collection of stops (for example, towns, stations,
More informationA constant approximation algorithm for the a priori traveling salesman problem
A constant approximation algorithm for the a priori traveling salesman problem David Shmoys 1 and Kunal Talwar 2 1 Cornell University, Ithaca NY 14853 shmoys@cs.cornell.edu 2 Microsoft Research, Silicon
More information2 Notation and Preliminaries
On Asymmetric TSP: Transformation to Symmetric TSP and Performance Bound Ratnesh Kumar Haomin Li epartment of Electrical Engineering University of Kentucky Lexington, KY 40506-0046 Abstract We show that
More informationNP-Completeness. Until now we have been designing algorithms for specific problems
NP-Completeness 1 Introduction Until now we have been designing algorithms for specific problems We have seen running times O(log n), O(n), O(n log n), O(n 2 ), O(n 3 )... We have also discussed lower
More informationEasy Problems vs. Hard Problems. CSE 421 Introduction to Algorithms Winter Is P a good definition of efficient? The class P
Easy Problems vs. Hard Problems CSE 421 Introduction to Algorithms Winter 2000 NP-Completeness (Chapter 11) Easy - problems whose worst case running time is bounded by some polynomial in the size of the
More informationJuly 18, Approximation Algorithms (Travelling Salesman Problem)
Approximation Algorithms (Travelling Salesman Problem) July 18, 2014 The travelling-salesman problem Problem: given complete, undirected graph G = (V, E) with non-negative integer cost c(u, v) for each
More informationNetwork Design and Game Theory Spring 2008 Lecture 6
Network Design and Game Theory Spring 2008 Lecture 6 Guest Lecturer: Aaron Archer Instructor: Mohammad T. Hajiaghayi Scribe: Fengming Wang March 3, 2008 1 Overview We study the Primal-dual, Lagrangian
More informationLP-relaxations for Tree Augmentation
LP-relaxations for Tree Augmentation Guy Kortsarz 1 and Zeev Nutov 2 1 Rutgers University Camden, NJ. guyk@camden.rutgers.edu 2 The Open University of Israel. nutov@openu.ac.il Abstract. In the Tree Augmentation
More informationarxiv: v2 [cs.ds] 4 Oct 2011
13 9-approximation for Graphic TSP Marcin Mucha Institute of Informatics, University of Warsaw, Poland mucha@mimuw.edu.pl arxiv:1108.1130v2 [cs.ds] 4 Oct 2011 Abstract The Travelling Salesman Problem is
More informationNew Integer Programming Formulations of the Generalized Travelling Salesman Problem
American Journal of Applied Sciences 4 (11): 932-937, 2007 ISSN 1546-9239 2007 Science Publications New Integer Programming Formulations of the Generalized Travelling Salesman Problem Petrica C. Pop Department
More informationOn the Integrality Ratio for the Asymmetric Traveling Salesman Problem
On the Integrality Ratio for the Asymmetric Traveling Salesman Problem Moses Charikar Dept. of Computer Science, Princeton University, 35 Olden St., Princeton, NJ 08540 email: moses@cs.princeton.edu Michel
More informationRelaxations and Bounds. 6.1 Optimality and Relaxations. Suppose that we are given an IP. z = max c T x : x X,
6 Relaxations and Bounds 6.1 Optimality and Relaxations Suppose that we are given an IP z = max c T x : x X, where X = x : Ax b,x Z n and a vector x X which is a candidate for optimality. Is there a way
More informationSection Notes 8. Integer Programming II. Applied Math 121. Week of April 5, expand your knowledge of big M s and logical constraints.
Section Notes 8 Integer Programming II Applied Math 121 Week of April 5, 2010 Goals for the week understand IP relaxations be able to determine the relative strength of formulations understand the branch
More informationPart III: Traveling salesman problems
Transportation Logistics Part III: Traveling salesman problems c R.F. Hartl, S.N. Parragh 1/282 Motivation Motivation Why do we study the TSP? c R.F. Hartl, S.N. Parragh 2/282 Motivation Motivation Why
More informationInteger Programming ISE 418. Lecture 8. Dr. Ted Ralphs
Integer Programming ISE 418 Lecture 8 Dr. Ted Ralphs ISE 418 Lecture 8 1 Reading for This Lecture Wolsey Chapter 2 Nemhauser and Wolsey Sections II.3.1, II.3.6, II.4.1, II.4.2, II.5.4 Duality for Mixed-Integer
More informationCombinatorial Optimization
Combinatorial Optimization Problem set 8: solutions 1. Fix constants a R and b > 1. For n N, let f(n) = n a and g(n) = b n. Prove that f(n) = o ( g(n) ). Solution. First we observe that g(n) 0 for all
More information15.081J/6.251J Introduction to Mathematical Programming. Lecture 24: Discrete Optimization
15.081J/6.251J Introduction to Mathematical Programming Lecture 24: Discrete Optimization 1 Outline Modeling with integer variables Slide 1 What is a good formulation? Theme: The Power of Formulations
More informationTravelling Salesman Problem
Travelling Salesman Problem Fabio Furini November 10th, 2014 Travelling Salesman Problem 1 Outline 1 Traveling Salesman Problem Separation Travelling Salesman Problem 2 (Asymmetric) Traveling Salesman
More informationEXACT ALGORITHMS FOR THE ATSP
EXACT ALGORITHMS FOR THE ATSP Branch-and-Bound Algorithms: Little-Murty-Sweeney-Karel (Operations Research, ); Bellmore-Malone (Operations Research, ); Garfinkel (Operations Research, ); Smith-Srinivasan-Thompson
More informationIterative Rounding and Relaxation
Iterative Rounding and Relaxation James Davis Department of Computer Science Rutgers University Camden jamesdav@camden.rutgers.edu March 11, 2010 James Davis (Rutgers Camden) Iterative Rounding 1 / 58
More informationA New Approximation Algorithm for the Asymmetric TSP with Triangle Inequality By Markus Bläser
A New Approximation Algorithm for the Asymmetric TSP with Triangle Inequality By Markus Bläser Presented By: Chris Standish chriss@cs.tamu.edu 23 November 2005 1 Outline Problem Definition Frieze s Generic
More informationDiscrete Optimization 2010 Lecture 12 TSP, SAT & Outlook
Discrete Optimization 2010 Lecture 12 TSP, SAT & Outlook Marc Uetz University of Twente m.uetz@utwente.nl Lecture 12: sheet 1 / 29 Marc Uetz Discrete Optimization Outline TSP Randomization Outlook 1 Approximation
More informationDiscrete Optimization 2010 Lecture 8 Lagrangian Relaxation / P, N P and co-n P
Discrete Optimization 2010 Lecture 8 Lagrangian Relaxation / P, N P and co-n P Marc Uetz University of Twente m.uetz@utwente.nl Lecture 8: sheet 1 / 32 Marc Uetz Discrete Optimization Outline 1 Lagrangian
More informationAlgorithm Design Strategies V
Algorithm Design Strategies V Joaquim Madeira Version 0.0 October 2016 U. Aveiro, October 2016 1 Overview The 0-1 Knapsack Problem Revisited The Fractional Knapsack Problem Greedy Algorithms Example Coin
More information3.4 Relaxations and bounds
3.4 Relaxations and bounds Consider a generic Discrete Optimization problem z = min{c(x) : x X} with an optimal solution x X. In general, the algorithms generate not only a decreasing sequence of upper
More informationCS599: Convex and Combinatorial Optimization Fall 2013 Lecture 17: Combinatorial Problems as Linear Programs III. Instructor: Shaddin Dughmi
CS599: Convex and Combinatorial Optimization Fall 2013 Lecture 17: Combinatorial Problems as Linear Programs III Instructor: Shaddin Dughmi Announcements Today: Spanning Trees and Flows Flexibility awarded
More informationarxiv: v2 [cs.ds] 2 Nov 2011
Improving Christofides Algorithm for thes-t Path TSP Hyung-Chan An Robert Kleinberg David B. Shmoys arxiv:1110.4604v [cs.ds] Nov 011 We present a deterministic Abstract 1+ ) 5 -approximation algorithm
More informationApproximation Algorithms for Asymmetric TSP by Decomposing Directed Regular Multigraphs
Approximation Algorithms for Asymmetric TSP by Decomposing Directed Regular Multigraphs Haim Kaplan Tel-Aviv University, Israel haimk@post.tau.ac.il Nira Shafrir Tel-Aviv University, Israel shafrirn@post.tau.ac.il
More informationDiscrete Optimization 2010 Lecture 12 TSP, SAT & Outlook
TSP Randomization Outlook Discrete Optimization 2010 Lecture 12 TSP, SAT & Outlook Marc Uetz University of Twente m.uetz@utwente.nl Lecture 12: sheet 1 / 29 Marc Uetz Discrete Optimization Outline TSP
More informationSeparating Simple Domino Parity Inequalities
Separating Simple Domino Parity Inequalities Lisa Fleischer Adam Letchford Andrea Lodi DRAFT: IPCO submission Abstract In IPCO 2002, Letchford and Lodi describe an algorithm for separating simple comb
More information4 Packing T-joins and T-cuts
4 Packing T-joins and T-cuts Introduction Graft: A graft consists of a connected graph G = (V, E) with a distinguished subset T V where T is even. T-cut: A T -cut of G is an edge-cut C which separates
More informationLower bounds on the size of semidefinite relaxations. David Steurer Cornell
Lower bounds on the size of semidefinite relaxations David Steurer Cornell James R. Lee Washington Prasad Raghavendra Berkeley Institute for Advanced Study, November 2015 overview of results unconditional
More informationCS675: Convex and Combinatorial Optimization Fall 2016 Combinatorial Problems as Linear and Convex Programs. Instructor: Shaddin Dughmi
CS675: Convex and Combinatorial Optimization Fall 2016 Combinatorial Problems as Linear and Convex Programs Instructor: Shaddin Dughmi Outline 1 Introduction 2 Shortest Path 3 Algorithms for Single-Source
More information16.1 Min-Cut as an LP
600.469 / 600.669 Approximation Algorithms Lecturer: Michael Dinitz Topic: LPs as Metrics: Min Cut and Multiway Cut Date: 4//5 Scribe: Gabriel Kaptchuk 6. Min-Cut as an LP We recall the basic definition
More informationarxiv: v1 [cs.dm] 20 Jun 2017
On the Integrality Gap of the Prize-Collecting Steiner Forest LP Jochen Könemann 1, Neil Olver 2, Kanstantsin Pashkovich 1, R. Ravi 3, Chaitanya Swamy 1, and Jens Vygen 4 arxiv:1706.06565v1 cs.dm 20 Jun
More informationPolynomial-Time Reductions
Reductions 1 Polynomial-Time Reductions Classify Problems According to Computational Requirements Q. Which problems will we be able to solve in practice? A working definition. [von Neumann 1953, Godel
More informationA Randomized Rounding Approach to the Traveling Salesman Problem
A Randomized Rounding Approach to the Traveling Salesman Problem Shayan Oveis Gharan Management Science and Engineering Stanford University Stanford, CA 94305 shayan@stanford.edu Amin Saberi + Management
More informationDiscrete Optimization 2010 Lecture 7 Introduction to Integer Programming
Discrete Optimization 2010 Lecture 7 Introduction to Integer Programming Marc Uetz University of Twente m.uetz@utwente.nl Lecture 8: sheet 1 / 32 Marc Uetz Discrete Optimization Outline 1 Intro: The Matching
More informationCS675: Convex and Combinatorial Optimization Fall 2014 Combinatorial Problems as Linear Programs. Instructor: Shaddin Dughmi
CS675: Convex and Combinatorial Optimization Fall 2014 Combinatorial Problems as Linear Programs Instructor: Shaddin Dughmi Outline 1 Introduction 2 Shortest Path 3 Algorithms for Single-Source Shortest
More informationc 2015 Society for Industrial and Applied Mathematics
SIAM J. DISCRETE MATH. Vol. 29, No. 2, pp. 915 939 c 2015 Society for Industrial and Applied Mathematics TSP TOURS IN CUBIC GRAPHS: BEYOND 4/3 JOSÉ CORREA, OMAR LARRÉ, AND JOSÉ A. SOTO Abstract. After
More informationFormulations and Algorithms for Minimum Connected Dominating Set Problems
Formulations and Algorithms for Minimum Connected Dominating Set Problems Abilio Lucena 1 Alexandre Salles da Cunha 2 Luidi G. Simonetti 3 1 Universidade Federal do Rio de Janeiro 2 Universidade Federal
More informationAlgorithms Design & Analysis. Approximation Algorithm
Algorithms Design & Analysis Approximation Algorithm Recap External memory model Merge sort Distribution sort 2 Today s Topics Hard problem Approximation algorithms Metric traveling salesman problem A
More informationImproving on the 1.5-Approximation of. a Smallest 2-Edge Connected Spanning Subgraph. September 25, Abstract
Improving on the 1.5-Approximation of a Smallest 2-Edge Connected Spanning Subgraph J. Cheriyan y A. Seb}o z Z. Szigeti x September 25, 1999 Abstract We give a 17 -approximation algorithm for the following
More informationThe Traveling Salesman Problem with Pickup and Delivery. A polyhedral approach. IBM Research - Australia. Irina Dumitrescu
Australia The Traveling Salesman Problem with Pickup and Delivery A polyhedral approach Irina Dumitrescu Jean-Francois Cordeau, Gilbert Laporte, Stefan Ropke The TSP with Pickup and Delivery (TSPPD) Given:
More informationNP-Completeness. ch34 Hewett. Problem. Tractable Intractable Non-computable computationally infeasible super poly-time alg. sol. E.g.
NP-Completeness ch34 Hewett Problem Tractable Intractable Non-computable computationally infeasible super poly-time alg. sol. E.g., O(2 n ) computationally feasible poly-time alg. sol. E.g., O(n k ) No
More informationOnline Supplement to A Branch and Price Algorithm for Solving the Hamiltonian p-median Problem
Online Supplement to A Branch and Price Algorithm for Solving the Hamiltonian p-median Problem Ahmed M. Marzouk, Erick Moreno-Centeno Industrial and Systems Engineering, Texas A&M University, College Station,
More informationLecture 6 January 21, 2013
UBC CPSC 536N: Sparse Approximations Winter 03 Prof. Nick Harvey Lecture 6 January, 03 Scribe: Zachary Drudi In the previous lecture, we discussed max flow problems. Today, we consider the Travelling Salesman
More informationScheduling and Optimization Course (MPRI)
MPRI Scheduling and optimization: lecture p. /6 Scheduling and Optimization Course (MPRI) Leo Liberti LIX, École Polytechnique, France MPRI Scheduling and optimization: lecture p. /6 Teachers Christoph
More informationWeek 8. 1 LP is easy: the Ellipsoid Method
Week 8 1 LP is easy: the Ellipsoid Method In 1979 Khachyan proved that LP is solvable in polynomial time by a method of shrinking ellipsoids. The running time is polynomial in the number of variables n,
More informationExercises NP-completeness
Exercises NP-completeness Exercise 1 Knapsack problem Consider the Knapsack problem. We have n items, each with weight a j (j = 1,..., n) and value c j (j = 1,..., n) and an integer B. All a j and c j
More informationA branch-and-cut algorithm for the Minimum Labeling Hamiltonian Cycle Problem and two variants
A branch-and-cut algorithm for the Minimum Labeling Hamiltonian Cycle Problem and two variants Nicolas Jozefowiez 1, Gilbert Laporte 2, Frédéric Semet 3 1. LAAS-CNRS, INSA, Université de Toulouse, Toulouse,
More informationLecture 15 - NP Completeness 1
CME 305: Discrete Mathematics and Algorithms Instructor: Professor Aaron Sidford (sidford@stanford.edu) February 29, 2018 Lecture 15 - NP Completeness 1 In the last lecture we discussed how to provide
More informationData Structures in Java
Data Structures in Java Lecture 21: Introduction to NP-Completeness 12/9/2015 Daniel Bauer Algorithms and Problem Solving Purpose of algorithms: find solutions to problems. Data Structures provide ways
More informationThe quest for finding Hamiltonian cycles
The quest for finding Hamiltonian cycles Giang Nguyen School of Mathematical Sciences University of Adelaide Travelling Salesman Problem Given a list of cities and distances between cities, what is the
More informationPart III: Traveling salesman problems
Transportation Logistics Part III: Traveling salesman problems c R.F. Hartl, S.N. Parragh 1/74 Motivation Motivation Why do we study the TSP? it easy to formulate it is a difficult problem many significant
More informationRESEARCH INSTITUTE FOR MATHEMATICAL SCIENCES RIMS Covering Cuts in Bridgeless Cubic Graphs. Sylvia BOYD, Satoru IWATA, Kenjiro TAKAZAWA
RIMS-1731 Covering Cuts in Bridgeless Cubic Graphs By Sylvia BOYD, Satoru IWATA, Kenjiro TAKAZAWA August 2011 RESEARCH INSTITUTE FOR MATHEMATICAL SCIENCES KYOTO UNIVERSITY, Kyoto, Japan Covering Cuts in
More informationApproximation Algorithms
Approximation Algorithms What do you do when a problem is NP-complete? or, when the polynomial time solution is impractically slow? assume input is random, do expected performance. Eg, Hamiltonian path
More informationlamsade 329 Décembre 2012 Laboratoire d Analyses et Modélisation de Systèmes pour l Aide à la Décision UMR 7243
lamsade Laboratoire d Analyses et Modélisation de Systèmes pour l Aide à la Décision UMR 7243 CAHIER DU LAMSADE 329 Décembre 2012 A note on Using T-joins to Approximate Solutions for MIN GRAPHIC K-PATH
More informationLecture 10 February 4, 2013
UBC CPSC 536N: Sparse Approximations Winter 2013 Prof Nick Harvey Lecture 10 February 4, 2013 Scribe: Alexandre Fréchette This lecture is about spanning trees and their polyhedral representation Throughout
More informationTractable & Intractable Problems
Tractable & Intractable Problems We will be looking at : What is a P and NP problem NP-Completeness The question of whether P=NP The Traveling Salesman problem again Programming and Data Structures 1 Polynomial
More informationCS/COE
CS/COE 1501 www.cs.pitt.edu/~nlf4/cs1501/ P vs NP But first, something completely different... Some computational problems are unsolvable No algorithm can be written that will always produce the correct
More informationNP and Computational Intractability
NP and Computational Intractability 1 Polynomial-Time Reduction Desiderata'. Suppose we could solve X in polynomial-time. What else could we solve in polynomial time? don't confuse with reduces from Reduction.
More informationSAT, Coloring, Hamiltonian Cycle, TSP
1 SAT, Coloring, Hamiltonian Cycle, TSP Slides by Carl Kingsford Apr. 28, 2014 Sects. 8.2, 8.7, 8.5 2 Boolean Formulas Boolean Formulas: Variables: x 1, x 2, x 3 (can be either true or false) Terms: t
More informationNow just show p k+1 small vs opt. Some machine gets k=m jobs of size at least p k+1 Can also think of it as a family of (1 + )-approximation algorithm
Review: relative approx. alg by comparison to lower bounds. TS with triangle inequality: A nongreedy algorithm, relating to a clever lower bound nd minimum spanning tree claim costs less than TS tour Double
More informationParity Polytopes and Binarization
Parity Polytopes and Binarization Dominik Ermel 1 and Matthias Walter 1 Otto-von-Guericke-Universität Magdeburg, dominik.ermel@st.ovgu.de RWTH Aachen University, walter@or.rwth-aachen.de March 8, 018 Abstract
More informationData Structures and Algorithms (CSCI 340)
University of Wisconsin Parkside Fall Semester 2008 Department of Computer Science Prof. Dr. F. Seutter Data Structures and Algorithms (CSCI 340) Homework Assignments The numbering of the problems refers
More informationDiscrete (and Continuous) Optimization WI4 131
Discrete (and Continuous) Optimization WI4 131 Kees Roos Technische Universiteit Delft Faculteit Electrotechniek, Wiskunde en Informatica Afdeling Informatie, Systemen en Algoritmiek e-mail: C.Roos@ewi.tudelft.nl
More informationThe Traveling Salesman Problem: Inequalities and Separation
The Traveling Salesman Problem: Inequalities and Separation Adam N. Letchford Department of Management Science, Lancaster University http://www.lancs.ac.uk/staff/letchfoa 1. The ILP formulation of the
More informationCuts & Metrics: Multiway Cut
Cuts & Metrics: Multiway Cut Barna Saha April 21, 2015 Cuts and Metrics Multiway Cut Probabilistic Approximation of Metrics by Tree Metric Cuts & Metrics Metric Space: A metric (V, d) on a set of vertices
More informationUnit 1A: Computational Complexity
Unit 1A: Computational Complexity Course contents: Computational complexity NP-completeness Algorithmic Paradigms Readings Chapters 3, 4, and 5 Unit 1A 1 O: Upper Bounding Function Def: f(n)= O(g(n)) if
More informationCable Trench Problem
Cable Trench Problem Matthew V Galati Ted K Ralphs Joseph C Hartman magh@lehigh.edu Department of Industrial and Systems Engineering Lehigh University, Bethlehem, PA Cable Trench Problem p.1 Cable Trench
More informationA stabilized column generation scheme for the traveling salesman subtour problem
Discrete Applied Mathematics 154 (2006) 2212 2238 www.elsevier.com/locate/dam A stabilized column generation scheme for the traveling salesman subtour problem Andreas Westerlund, Maud Göthe-Lundgren, Torbjörn
More informationInteger Linear Programming (ILP)
Integer Linear Programming (ILP) Zdeněk Hanzálek, Přemysl Šůcha hanzalek@fel.cvut.cz CTU in Prague March 8, 2017 Z. Hanzálek (CTU) Integer Linear Programming (ILP) March 8, 2017 1 / 43 Table of contents
More informationWorst Case Performance of Approximation Algorithm for Asymmetric TSP
Worst Case Performance of Approximation Algorithm for Asymmetric TSP Anna Palbom Department of Numerical Analysis and Computer Science Royal Institute of Technology S-100 44 Stockholm, Sweden E-mail: annap@nada.kth.se
More informationSome Algebra Problems (Algorithmic) CSE 417 Introduction to Algorithms Winter Some Problems. A Brief History of Ideas
Some Algebra Problems (Algorithmic) CSE 417 Introduction to Algorithms Winter 2006 NP-Completeness (Chapter 8) Given positive integers a, b, c Question 1: does there exist a positive integer x such that
More informationWeek Cuts, Branch & Bound, and Lagrangean Relaxation
Week 11 1 Integer Linear Programming This week we will discuss solution methods for solving integer linear programming problems. I will skip the part on complexity theory, Section 11.8, although this is
More informationORIE 6334 Spectral Graph Theory November 22, Lecture 25
ORIE 64 Spectral Graph Theory November 22, 206 Lecture 25 Lecturer: David P. Williamson Scribe: Pu Yang In the remaining three lectures, we will cover a prominent result by Arora, Rao, and Vazirani for
More informationWhat Computers Can Compute (Approximately) David P. Williamson TU Chemnitz 9 June 2011
What Computers Can Compute (Approximately) David P. Williamson TU Chemnitz 9 June 2011 Outline The 1930s-40s: What can computers compute? The 1960s-70s: What can computers compute efficiently? The 1990s-:
More informationLinear Programming. Scheduling problems
Linear Programming Scheduling problems Linear programming (LP) ( )., 1, for 0 min 1 1 1 1 1 11 1 1 n i x b x a x a b x a x a x c x c x z i m n mn m n n n n! = + + + + + + = Extreme points x ={x 1,,x n
More informationThe 2-valued case of makespan minimization with assignment constraints
The 2-valued case of maespan minimization with assignment constraints Stavros G. Kolliopoulos Yannis Moysoglou Abstract We consider the following special case of minimizing maespan. A set of jobs J and
More informationComputational Integer Programming. Lecture 2: Modeling and Formulation. Dr. Ted Ralphs
Computational Integer Programming Lecture 2: Modeling and Formulation Dr. Ted Ralphs Computational MILP Lecture 2 1 Reading for This Lecture N&W Sections I.1.1-I.1.6 Wolsey Chapter 1 CCZ Chapter 2 Computational
More informationTraveling Salesman. Given G = (V,V V), find simple cycle C = (v 1,v 2,...,v n,v 1 ) such that n = V and (u,v) C d(u,v) is minimized.
Sanders/van Stee: Approximations- und Online-Algorithmen 1 Traveling Salesman Given G = (V,V V), find simple cycle C = (v 1,v 2,...,v n,v 1 ) such that n = V and (u,v) C d(u,v) is minimized. Sanders/van
More informationTraveling Salesman Problem
Traveling Salesman Problem Zdeněk Hanzálek hanzalek@fel.cvut.cz CTU in Prague April 17, 2017 Z. Hanzálek (CTU) Traveling Salesman Problem April 17, 2017 1 / 33 1 Content 2 Solved TSP instances in pictures
More informationChain-constrained spanning trees
Math. Program., Ser. A (2018) 167:293 314 https://doi.org/10.1007/s10107-017-1126-7 FULL LENGTH PAPER Chain-constrained spanning trees Neil Olver 1,2 Rico Zenklusen 3 Received: 2 November 2015 / Accepted:
More informationThe Ellipsoid (Kachiyan) Method
Yinyu Ye, MS&E, Stanford MS&E310 Lecture Note: Ellipsoid Method 1 The Ellipsoid (Kachiyan) Method Yinyu Ye Department of Management Science and Engineering Stanford University Stanford, CA 94305, U.S.A.
More informationNot Every GTSP Facet Induces an STSP Facet
Not Every GTSP Facet Induces an STSP Facet Marcus Oswald, Gerhard Reinelt, and Dirk Oliver Theis Institute of Computer Science University of Heidelberg Im Neuenheimer Feld 38, 9120 Heidelberg, Germany
More informationSome Open Problems in Approximation Algorithms
Some Open Problems in Approximation Algorithms David P. Williamson School of Operations Research and Information Engineering Cornell University August 18, 2011 APPROX 2011 David P. Williamson (Cornell
More information